N_roberta_sst5_padding70model

This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.5445
• Accuracy: 0.5421

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3521	1.0	534	1.3056	0.4149
1.0582	2.0	1068	1.0707	0.5281
0.8976	3.0	1602	1.0814	0.5380
0.7673	4.0	2136	1.1237	0.5602
0.6478	5.0	2670	1.2495	0.5439
0.5279	6.0	3204	1.3878	0.5448
0.4228	7.0	3738	1.5626	0.5357
0.3393	8.0	4272	1.7503	0.5195
0.282	9.0	4806	1.8795	0.5443
0.2473	10.0	5340	2.1451	0.5312
0.2186	11.0	5874	2.3606	0.5380
0.1938	12.0	6408	2.8212	0.5353
0.1642	13.0	6942	3.0636	0.5371
0.1602	14.0	7476	3.0900	0.5421
0.116	15.0	8010	3.2026	0.5471
0.0971	16.0	8544	3.2785	0.5376
0.0667	17.0	9078	3.3938	0.5448
0.07	18.0	9612	3.5817	0.5326
0.0633	19.0	10146	3.4982	0.5475
0.0608	20.0	10680	3.5445	0.5421

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3